Automotive Case Study
Onepoint helps automotive data provider in Europe bring consistency in handling data on vehicle specification, parts and pricing across different geographies.
Our client is the leading European provider of automotive data, insight and business intelligence services to global manufacturers, insurers, dealers, bodyshops and fleet and finance companies.
Business Drivers
- Data being the primary ingredient of their product and service offerings it was vital for them to ensure it was consistent, timely and reliable.
- To reduce the multiple systems with one common system.
- To reduce the time to produce output files for end customers.
- To stop loss of business renewal and opportunities.
- To support existing file and data formats but also enable faster generation of new and more flexible formats for clients.
- To offer more flexibility to their clients by extending the currently supported file and data formats to support.
The Challenge
- Firstly we were tasked to look at various solution options for the management of the vehicle master data.
- There was a need for common reference data that could be managed collaboratively including quality control.
- We were required to develop part of their strategic data model to help bring consistency in the way vehicle specification, parts and pricing data would be handled in a standardized manner across their operating countries. The client, through the combination of data collection, market analysis and mathematical models creates reliable, decision-critical information.
- We then requested to produce a Proof of concept (PoC) that demonstrated the ability to support existing data extract output formats to allow use of existing downstream systems and also so that end clients do not see any changes in the data they received. The information is available across Europe and makes a measurable contribution to increasing the profitability of the client’s customers.
The Solution
- Talend MDM (master data management) was selected and an extensive proof of concept (PoC) was delivered during the period March 2012 – November 2012. An option to develop a bespoke MDM solution was deemed to be too expensive.
- The new data model was delivered to the client having assessed 800+ data attributes and mapped that to a new data model comprising of 90+ entities.
- Logical data model created in the Sybase Power Designer tool.
- Physical data model created in Talend MDM with support for additional metadata required for permission control, workflow support, translations and rules (consistency, plausibility, combination, and price).
- We were not provided with a comprehensive list of NFRs (non-functional requirements) and so we used the following for reference
– Per country
– Number of edit requests per second: 3
– Number of searches (finding entity buy combination of attributes) per second: 3
– Number of read requests (retrieving all attributes of the entity) per second: 3
– Maximum latency under above load (seconds) – how long each request takes: 0.15
– Total number of entries (rows in all tables) per country: 10M.
The Outcome
The PoC was used to confirm that Talend MDM could be used as a backend store for shared (central) and local (to each country) data. The PoC high level logical architecture is summarized in the diagram above.
The newly implemented Talend technology brings with it additional benefits to ensure future business improvements to support Big data integration and draw insights from these data (Business Intelligence & Predictive Analytics).
